Vol. 41 2015 No. 2 DOI: 10.5277/epe150202
WIOLETTA ROGULA-KOZŁOWSKA1
CHEMICAL COMPOSITION AND MASS CLOSURE
OF AMBIENT PARTICULATE MATTER AT A CROSSROADS
AND A HIGHWAY IN KATOWICE, POLAND
The concentration, chemical composition, and mass closure of various fractions of ambient partic-ulate matter (PM) were analyzed at crossroads and at a highway in Katowice (Poland). It was shown that at both sites organic carbon can constitute even 57% of the fine PM mass, about 20% of the fine PM mass can originate from the photochemical transformations of inorganic gaseous precursors, and that the coarse PM was mainly the organic matter (up to 39%) and crustal matter (up to 24%). Traffic emissions in Katowice can affect the formation of secondary aerosol (organic and inorganic), the con-tributions to PM and ambient concentrations of soil matter, NaCl, and trace elements. At the high-way, the greatest impact on the concentrations and chemical composition of fine particles was proba-bly due to exhaust emissions. At the crossroads, in the center of the city, the non-exhaust traffic emissions probably affected the coarse PM.
1. INTRODUCTION
It is quite obvious that the growth of the road traffic causes the growth of the con-centrations of ambient particulate matter (PM) and increases the health hazard from it [1, 2]. Within urbanized regions, at trafficked sites, the effects of the traffic-related PM and PM precursors are reflected in the specific size distribution of the particles and the chemical composition of PM [3].
The chemical composition of PM has been studied worldwide for many years, in general to assess the PM effects on human health and on the environment, and to es-tablish methods for tracing the PM origin [4–7]. The origin of PM can be determined quite accurately by apportioning the gravimetrically determined PM mass between the groups of the analytically found PM components (PM mass closure) [e.g. 8–12]. The _________________________
1Institute of Environmental Engineering of the Polish Academy of Sciences, ul. M. Skłodowskiej-
PM components are most often grouped into elemental carbon (EC), organic matter (OM), secondary inorganic aerosol (SIA), crustal or mineral matter (CM or MM), sea salt or sodium chloride (SS or NaCl), trace or other elements (TE or OE), and the pos-sibly small rest-unidentified matter (UM).
The first investigations of the chemical composition and mass closure of the par-ticular PM fractions in Poland suggest that in a typical Upper Silesian city (Zabrze) over 80% of PM2.5 (throughout the whole paper, x – y in PMx–y denotes the interval of
the aerodynamic diameters [µm] of the particles in PMx–y; PMx stands for PM0.03–x and
50% of PM2.5–10 may be directly or indirectly linked to combustion of fuels, and that
almost 50% of PM1 may be secondary (inorganic or organic) aerosol from
transfor-mations of gaseous precursors [11, 12].
The goal of the presented work was to analyze the concentration, chemical com-position, and mass closure of each of PM1, PM1–2.5, PM2.5–10 and PM10–40 at two sites
strongly affected by traffic emissions and located within a typical urban area in the Upper Silesia Agglomeration. There is a need for the investigations of the effects of traffic emissions on the air quality in Poland, and it stems from the outdated road in-frastructure and the dynamic vehicle fleet development during the last two decades.
2. METHOD
PM was sampled at two sites in Katowice (Poland). One of the measuring points (HW) was located close to the A4 highway, about 1.5 km south of the city center. The other point (CR) was located close to two busy crossroads and the greatest traffic cir-cle in Katowice (Fig. 1).
At HW, PM was sampled from 13.03.2012 to 19.06.2012; at CR-from 20.06.22012 to 03.09.2012. There were 9 samplings at each sampling point; a single sampling lasted about one week (142 to 173 h at HW, 123 to 145 h at CR).
PM was sampled with a thirteen-stage DEKATI low pressure impactor (DLPI, flow rate 30 dm3/min). The impactor takes samples of 13 PM fractions (onto substrate
filters): PM0.03–0.06, PM0.06–0.108, PM0.108–0.17, PM0.17–0.26, PM0.26–0.4, PM0.4–0.65, PM0.65–1,
PM1–1.6, PM1.6–2.5, PM2.5–4.4, PM4.4–6.8, PM6.8–10, PM10–40. The samples of the fractions
PM1, PM1–2.5, PM2.5–10 and PM10–40 were the summed samples of all their sub-fractions,
their ambient concentrations were the sums of their sub-fraction concentrations. The PM samples were taken onto quartz filters (Whatman, QMA, 25 mm, CAT No. 1851-025) and nylon filters (Whatman, nylon membrane filters 0.2 µm, 25 mm, Cat No. 7402-002), alternating between samplings. Before and after exposing, all the filters were conditioned in a weighing room (48 h, relative air humidity 45±5%, air temperature 20±2 °C) and weighed twice, with 24 h period between, on a RADWAG microbalance (resolution 1 µg).
The PM sampled onto quartz filters (4 one-week samples at each point) was ana-lyzed for organic (OC) and elemental (EC) carbon. PM on the nylon filters (5 one- -week samples at each point) was analyzed first for its elemental composition (Al, Si, Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, As, Se, Br, Rb, Sr, Mo, Ag, Cd, Sb, Te, Ba and Pb) by means of a nondestructive method based on energy dispersive X-ray fluores-cence (EDXRF), then the samples were extracted with water and the concentrations of main water-soluble ions (Cl–, 2
4 3 4
SO , NO , Na , NH , K+, Ca2+, and Mg2+) were
de-termined in the extracts.
The parameters of the equipment (Sunset Laboratory carbon analyzer, Epsilon 5 EDXRF spectrometer, Herisau Metrohm AG ion chromatograph), the detailed descrip-tions of the analytical procedures, and the results of the validation of the method have been described elsewhere [11, 12]. The arithmetic averages of the weakly concentra-tions of PM and its components were assumed as the average ambient concentraconcentra-tions in the sampling periods (Table 1).
The PM mass reconstruction (mass closure) based on the PM component division into elemental carbon (EC), organic matter (OM), secondary inorganic aerosol (SIA), the group comprising Cl– and Na+ (NaCl), crustal matter (CM), trace elements (TE),
and unidentified matter (UM). The collective masses of the groups were computed from the analytically determined or calculated masses of their components. The sum of the masses of EC, OM, SIA, NaCl, CM, and TE was compared with the gravimetri-cally determined mass of PM. The mass closure was checked for each of PM1, PM1–2.5,
PM2.5–10 and PM10–40 and each sampling site.
The mass of EC was assumed to be the analytically determined mass [EC]A of
elemental carbon: [EC] = [EC]A. The mass of OM (OM, all PM bound organic
[OM] = 1.4[OC]A. SIA consisted of SO , NO , and NH24 3 4 and [SIA] = [SO ]24 A 3 A + [NO ] 4 A + [NH ] ; [NaCl] = [Cl–] A + [Na+]A.
CM and TE consisted of the elements listed in Table 1; the contents of CM and TE depended on the PM fraction and the measuring point. The elements were divided into two parts based on the analysis of their enrichment factors (EF, Table 2). The method for computing the enrichment factors is given in [12, 13]. For any element, EF is not lower than 0; the closer its value to 1, the smaller the anthropogenic effect on the ele-ment amount in the air.
The elements with EF ≤ 20 were considered crustal (bolds in Table 2), and CM was
assumed to include 2
3 CO , SiO
2, Al2O3, Mg2+, Ca2+, K2O, FeO and Fe2O3 (without FeO
and Fe2O3 for PM1–2.5 at HW and PM1 at CR), Rb and Sr (except Sr in PM1 at both HW
and CR) and Ba, (only in PM2.5–10 and PM10–40 at CR). The masses of the CM
compo-nents were computed stoichiometrically from the analytically determined masses of their component elements assuming that Fe is equally distributed between FeO and Fe2O3
and; the 2
3
CO mass was calculated from the masses of Ca2+ and Mg2+ [14].
The elements with EF > 20 were in TE: Sc, Ti, V, Cr, Mn, Fe (Fe only in PM1–2.5
at HW and in PM1 at CR), Co, Ni, Cu, Zn, As, Se, Br, Rb and Sr (Rb and Sr only in
PM1 at both HW and CR), Mo, Ag, Cd, Sb, Te, Ba (except Ba in PM2.5–10 and PM10–40
at CR). The PM bound compounds of these elements were not identified. Thus, the mass [TE] of TE was assumed to be the sum of all the analytically determined masses of the elements from TE.
The mass of unidentified matter [UM] was the deficient mass; it was the differ-ence between the gravimetrically determined mass of PM and [SIA] + [EC] + [OM] + [NaCl] + [CM] + [TE].
3. RESULTS AND DISCUSSION
3.1. CONCENTRATIONS OF PM AT THE SAMPLING POINTS
The concentration2 of total PM (summed concentrations of PM
1, PM1–2.5, PM2.5–10,
and PM10–40) at HW was higher than at CR (Table 1). The PM1 and PM1–2.5
concentra-tions at HW were very different from those at CR. The concentraconcentra-tions of PM2.5–10 and
PM10–40 were a bit higher at CR than at HW. The mass shares (percentages) of fine
particles (PM1 and PM1–2.5) in total PM were lower at CR than at HW (the former by 5
the latter by 8%).
_________________________
2In the paper, concentration, or ambient concentration denote the average concentrations in the
T a b l e 1 Sampling period averages of concentrations of PM, OC, EC [µg/m3],
water soluble ions, and the remaining components [ng/m3] of PM
1, PM1–2.5, PM2.5–10
and PM10–40 at the highway (HW) and the crossroads (CR) in Katowice
Species HW (13.03.2012–19.06.2012) CR (20.06.2012–03.09.2012) PM1 PM1–2.5 PM2.5–10 PM10–40 PM1 PM1–2.5 PM2.5–10 PM10–40 PM 18.75 5.20 4.72 1.21 12.14 3.25 4.98 1.67 OC 5.57 1.77 1.29 0.25 4.93 1.22 1.39 0.36 EC 0.89 0.24 0.46 0.07 0.52 0.13 0.29 0.10 Na+ 38.33 29.48 71.63 16.17 50.63 31.93 41.58 7.79 NH4+ 938.63 234.66 39.22 7.51 608.12 67.60 30.95 8.43 K+ 56.31 17.40 46.23 11.78 67.57 9.71 11.58 2.27 Ca2+ 24.16 2.76 45.28 10.92 35.28 17.63 71.07 41.19 Mg2+ 2.57 < DL < DL < DL 1.41 1.71 7.03 1.53 Cl– 459.97 125.63 177.01 50.80 223.58 66.36 125.68 37.02 3 NO 1113.63 362.12 206.46 32.95 458.16 172.51 254.96 54.16 2 4 SO 1437.18 360.41 166.67 46.01 1481.14 236.17 172.46 59.20 Al 12.41 17.20 43.75 11.05 8.94 34.11 61.49 15.20 Si 22.98 72.82 167.01 46.39 52.12 124.23 229.97 63.12 Sc 3.67 6.76 15.73 4.77 4.30 7.90 23.84 7.24 Ti 92.67 33.50 48.66 14.97 104.66 32.98 52.26 15.74 V 12.26 4.94 6.91 1.99 13.79 4.55 7.17 2.13 Cr 4.48 1.87 2.39 0.75 6.73 2.05 2.86 0.88 Mn 28.56 12.72 18.96 5.19 33.80 11.88 20.09 6.44 Fe 91.08 223.72 264.52 42.74 100.84 157.69 256.52 55.52 Co 2.14 1.10 1.34 0.35 1.66 0.47 0.66 0.19 Ni 0.50 0.29 0.36 0.07 0.52 0.21 0.33 0.08 Cu 7.08 7.33 6.72 0.84 7.61 3.40 4.08 0.86 Zn 38.42 24.96 14.52 3.84 35.86 34.34 28.35 6.51 As 6.62 3.97 1.60 0.42 6.71 4.57 2.87 0.65 Se 0.39 0.13 0.00 0.01 0.27 0.05 <DL <DL Br 6.46 1.41 0.51 0.16 3.51 0.23 0.18 0.08 Rb 0.77 0.20 0.11 0.02 0.49 0.09 0.15 0.03 Sr 1.75 1.03 1.55 0.42 2.07 0.86 1.28 0.32 Mo 1.04 0.56 0.77 0.21 1.31 0.69 0.85 0.28 Ag 1.98 0.48 0.76 0.28 2.18 0.54 0.84 0.20 Cd 2.98 1.16 1.16 0.35 3.19 0.92 1.22 0.38 Sb 51.23 17.05 22.26 6.97 55.90 14.71 22.59 7.15 Te 1.64 0.44 0.52 0.14 1.82 0.57 0.93 0.27 Ba 12.55 8.04 9.83 2.11 13.04 6.24 8.41 2.52 Pb 18.45 11.68 4.60 1.19 18.76 13.86 8.56 1.95
Almost all primary particles (mainly soot) from combustion of oil derivatives in car engines are fine PM; the majority have aerodynamic diameters not greater than 0.5 µm [16, 17]. Thus, the greatest impact on the fine particles concentrations at HW was probably due to exhaust emissions. At CR in the Katowice center, the re-suspended PM lifted by vehicles and pedestrians, an important PM source in the peri-ods of low precipitation, probably affected the PM concentrations. Also the particles, mainly coarse [3, 18], from corrosion or abrasive wear of vehicles and road surface (braking, stopping, moving off) might have a stronger effect on PM at CR than at HW.
T a b l e 2 Sampling period averages of the enrichment factors (EFa) of the elements from PM
1,
PM1–2.5, PM2.5–10 and PM10–40, at a crossroads (CR) and a highway (HW) in Katowice
Species [ppm] UC HW (13.03.2012–19.06.2012) CR (20.06.2012–03.09.2012) PM1 PM1–2.5 PM2.5–10 PM10–40 PM1 PM1–2.5 PM2.5–10 PM10–40 Mg2+ 13510 1 0 0 0 1 0 1 1 Al 77440 1 1 1 1 1 1 1 1 Si 303480 1 1 1 1 2 1 1 1 K+ 28650 12 3 3 3 20 1 1 0 Ca2+ 29450 5 0 3 3 10 1 3 7 Sc 7 3273 4345 3978 4775 5325 2562 4289 5270 Ti 3117 186 48 28 34 291 24 21 26 V 53 1444 420 231 264 2254 195 170 205 Cr 35 798 241 121 149 1666 133 103 128 Mn 527 338 109 64 69 556 51 48 62 Fe 30890 18 33 15 10 28 12 11 9 Co 11.6 1153 428 204 213 1237 92 72 84 Ni 18.6 167 70 34 25 241 26 23 22 Cu 14.3 3092 2307 832 410 4607 540 359 307 Zn 52 4612 2161 494 517 5974 1499 687 638 As 2 20654 8946 1417 1457 29059 5190 1807 1650 Se 0.1 29609 6888 72 572 28169 1366 0 0 Br 1.6 25182 3961 566 680 18999 331 140 246 Rb 110 43 8 2 1 39 2 2 2 Sr 316 35 15 9 9 57 6 5 5 Mo 1.4 4619 1804 968 1030 8098 1122 763 1021 Ag 0.1 224325 39581 24560 35260 343123 22186 19121 18894 Cd 0.1 182216 51288 20191 24363 271235 20489 15017 19180 Sb 0.3 1031550 247525 127084 157726 1562017 107699 91768 117502 Te – – – – – – – – – Ba 668 117 54 26 22 169 21 16 19 Pb 17 6776 3093 479 492 9559 1851 634 583
aEF are computed relative to the concentration of Al, a marker element for the Earth crust (EF Al = 1).
The differences in the PM concentrations between CR and HW reflect the differ-ences in the emissions of the PM and PM precursors between the two measuring peri-ods in Katowice. The monthly concentrations of PM2.5 and PM10 taken from the State
Environmental Monitoring Program station in Katowice (urban background) in 2012 are presented in Fig. 2 (http://stacje.katowice.pios.gov.pl/monitoring/).
Fig. 2. Monthly concentrations of PM2.5 and PM10 at the urban background point
in Katowice in 2012 (http://stacje.katowice.pios.gov.pl/monitoring/)
In the holiday period (July–August), they were lower than in the rest of the year, especially those of PM2.5 that are kept high by municipal and power plant emissions all
over Poland not only in winter but also in March–June, the measuring period at HW. In the summer holiday period, vehicle and pedestrian traffic decreases, also municipal emissions decrease because of lower household demands for energy and fuels. The average PM2.5 concentration (arithmetic mean of the monthly June, July, and August
concentrations) at the urban background point in Katowice was equal to the average PM2.5 concentration at CR (15 µg/m3); the average PM10 concentration at the urban
background point (40 µg/m3) was even higher than at CR.
3.2. CHEMICAL COMPOSITION AND MASS CLOSURE OF PM1, PM1–2.5, PM2.5–10 AND PM10–40
At both sampling points, most of the mass of each PM fraction was in carbona-ceous matter (OM and EC, Table 3, Fig. 3). The ambient OC concentrations at HW and at CR resembled the concentrations in other regions of Upper Silesia or traffic affected sites in Europe [11, 19]. The EC concentrations were much lower in Katowi-ce, especially at CR (summer). In urban, and sometimes in rural regions of southern Poland, the winter EC concentrations are significantly higher than elsewhere in Eu-rope, sometimes even than at traffic affected sites [11, 12, 19]. In Katowice, the EC concentrations are affected more by municipal emissions than by road traffic – unlike elsewhere in Europe [6]; it is confirmed by the EC distribution among the fractions: in Katowice, EC did not concentrate only in fine PM. Except for almost 10% EC
contri-bution to PM2.5–10 at HW, all other fractional mass contributions were about 5%
(Table 3). The mass of EC in the coarsest PM, PM10–40, was 5.5% at HW and 5.8% at
CR; in PM1 – 4.8% and 4%. Thus, the coarse particles (PM2.5–10 and PM10–40) were
very rich in EC.
Fig. 3. Concentrations of organic matter (OM), elemental carbon (EC),
secondary inorganic aerosol (SIA), sum of sodium and chlorides (NaCl), crustal matter (CM), trace elements (TE), and unidentified matter (UM) related to PM1, PM1–2.5, PM2.5–10 and PM10–40
at the highway (HW) and at the crossroads (CR) in Katowice
T a b l e 3 Chemical mass closure of PM1, PM1–2.5, PM2.5–10 and PM10–40
at the highway (HW) and at the crossroads (CR) in Katowice [%]
Species HW (13.03.2012–19.06.2012) CR (20.06.2012–03.09.2012) PM1 PM1–2.5 PM2.5–10 PM10–40 PM1 PM1–2.5 PM2.5–10 PM10–40 EC 4.8 4.6 9.7 5.5 4.3 4.0 5.8 5.8 OM 41.6 47.6 38.3 28.6 56.8 52.7 39.1 30.0 SIA 18.6 18.4 8.7 7.2 21.0 14.7 9.2 7.3 NaCl 2.7 3.0 5.3 5.6 2.3 3.0 3.4 2.7 CM 1.8 10.0 20.6 18.3 2.5 18.7 23.6 21.0 TE 1.6 2.7 3.3 3.7 2.6 4.3 3.7 3.2 UM 29.0 13.7 14.1 31.2 10.6 2.6 15.1 30.1
OM – organic matter, EC – elemental carbon, SIA – secondary inorganic aero-sol, NaCl – sum of sodium and chlorides, CM – crustal matter, TE – trace ele-ments, UM – unidentified matter.
In Katowice, in summer, the municipal emissions can contribute much to ambient EC. All over Silesia, inefficient coal-fired stoves or boilers emit big agglomerates of soot from incomplete combustion of coal during the whole year [19]; also the Polish
vehicle fleet, in general obsolete cars with ineffective exhaust control systems, emits agglomerated soot. In general, at both sites, OM contributed more to fine than to coarse PM; at CR the OM mass contribution to PM1 was 56.8%. In each fraction, the
OM mass was greater at CR than at HW (Table 3); the differences amounted from 0.8% (PM2.5–10) to 15.2% (PM1). It was due to the PM concentrations, lower at CR
than at HW, because the OC ambient concentrations at CR were lower than at HW (Table 1).
The mass contributions of secondary (OCsec) and primary (OCprim) organic carbon
to organic carbon in PM1, PM1–2.5, PM2.5–10 and PM10–40 at both sampling points are
presented in Fig. 4.
Fig. 4. Shares [%] of secondary organic carbon (OCsec) and primary organic carbon (OCprim)
in organic carbon (OC) bound to PM1, PM1–2.5, PM2.5–10, PM10–40, and TSP
at the highway (HW) and at the crossroads (CR)
For the fraction f at the point p, they were computed from the following formulas [20]:
A
A A min sec A OC OC OC 100% EC p s f f s p p EC (1)
prim sec OC f 100 OC f p p (2) where:
OCsec
fpis the mass contribution of OCsec to the fraction f bound OC at the point p,
prim
OC f
p
OC fAp is the analytically determined mass of the fraction f bound OC at the point p,
EC Af
p is the analytically determined mass of the fraction f bound EC at the point p,
OC Asis the analytically determined mass of OC from the sample s,
EC Asis the analytically determined mass of EC from the sample s,
OCA/ EC A min
p
s s
is the minimum
OC As / EC Asover all the samples at the point p,
OCs / EC s
minHW A A = 1.034,
OC A/ ECA min
CR s s = 2.793.OCsec (and secondary OM) arises from condensation of gaseous organic
com-pounds and photochemical transformations of volatile organic comcom-pounds [20, 21]. Both sampling points were within the effect of road traffic causing high concentrations of nitric oxides, ozone, and hydrocarbons favorable for the formation of OCsec. The
insolation was also high at both sites (more intense during measurements at CR, (http://stacje.katowice.pios.gov.pl/monitoring/). Consequently, at both sampling points, OCsec prevailed in PM1 and PM1–2.5 bound OC. The share of OCsec in PM2.5–10 and
PM10–40 bound OC was not greater than 40% at CR and was almost 70% at HW (Fig.
4). At HW, it might be due to the high percent of EC in coarse PM (condensation of gaseous organic compounds on the vast surfaces of EC particles); also the share of biological matter in OC (plant debris from lawn, pollen) was greater at CR than at HW.
Secondary inorganic aerosol (SIA), the PM part coming from the photochemical transformations of inorganic gaseous precursors, was about 16.5% of the total PM mass at both sites; it was 18.6% of the PM1 mass at HW and 21% at CR; its parts in
PM2.5–10 and PM10–40 were small.
T a b l e 4 Ambient concentrations [µg/m3] of ammonium sulfate and ammonium nitrate
bound to PM1, PM1–2.5, PM2.5–10 and PM10–40, and neutralization ratios (NR)
for these fractions at the highway (HW) and at the crossroads (CR) in Katowice
Species PMHW (13.03.2012–19.06.2012) CR (20.06.2012–03.09.2012) 1 PM1–2.5 PM2.5–10 PM10–40 PM1 PM1–2.5 PM2.5–10 PM10–40 NR 1.08 0.96 0.32 0.28 0.88 0.48 0.22 0.22 (NH4)2SO4 1.98 0.50 0.14 0.03 2.23 0.25 0.11 0.03 ex-NH4+ 0.40 0.10 0.00 0.00 0.00 0.00 0.00 0.00 NH4NO3 1.79 0.45 0.00 0.00 0.00 0.00 0.00 0.00
The capability of ambient ammonium (NH )4 to neutralize ambient sulfates 2 4 (SO ) and nitrates (NO )3 can be expressed as the neutralization ratio NR, i.e. the proportion of the concentration of NH4 [neq/m3] to the collective concentration of 2
4 SO and
3 NO [neq/m3] (Table 4). NR was computed for each PM fraction and each sampling point.
The amount of NH4+ in a fraction with NR ≥ 1 suffices for the total neutralization
of the sulfuric (H2SO4) and nitric (HNO3) acids from this fraction. The mass
[(NH4)2SO4] of (NH4)2SO4 in a fraction can be computed from the equation:
[(NH4)2SO4] = 1.38[SO42–]A (3)
The mass [ex-NH4+] of the fraction bound NH4+ left after its reaction with SO42– and
the mass [NH4NO3] of the fraction bound NH4NO3 were computed from:
[ex-NH4+] = [NH4+]A – 0.27[(NH4)2SO4] (4)
[NH4NO3] = 4.44[ex-NH4+] (5)
NH4+ in the fraction with NR < 1 can neutralize only a part of SO42– from this
frac-tion. It is, (NH4)2SO4 arises and there is too little NH4+ for NH4NO3 to be formed. The
mass of (NH4)2SO4 from these fractions was computed from:
[(NH4)2SO4] = 3.67[NH4+]A (6)
Table 4 presents the results of stoichiometric computations applied to the ambient concentrations of sulfates (SO42–), nitrates (NO3–), and ammonium (NH4+). The
compu-tations, despite their simplicity, reveal clearly some regularity: NH4NO3 did not occur
in SIA in any PM fraction at CR. At HW, the ambient concentrations of nitrates, much
greater than at CR, could be neutralized by NH4+. So, at HW, ammonium sulfate
((NH4)2SO4) and ammonium nitrate (NH4NO3) were present in fine PM bound SIA.
Because the highway is the only pollution source that could cause high ambient con-centrations of nitric oxides or ammonia, the precursors of NH4NO3, at HW, the
com-position of PM1 and PM1–2.5 bound SIA must have been formed by traffic. In the
holi-day season, the traffic effects at CR were weaker – the concentrations of PM1 and
PM1–2.5 bound NO3– and NH4+ were much lower than at HW (Table 1).
Small NR for PM2.5–10 and PM10–40 at both HW and CR mean the lack of NH4NO3
in these fractions; probably, at both the sites, sulfates and nitrates occurred as CaSO4,
NaNO3, or (Na)2SO4 in coarse PM.
The CM and NaCl mass shares in PM fractions were noticeable at both HW and CR (Table 3, Fig. 3). The CM shares in PM2.5–10 and PM10–40 were about 20% at both
frac-tion coming mainly from combusfrac-tion (high OM + EC) and from the transformafrac-tions of gaseous PM precursors (SIA) did not exceed 2.5% at both points. The shares of NaCl in fine PM did not exceed 3% at both points, in coarse PM at HW it was almost 5.6%.
At HW, the salt from winter road deicing, accumulated during winter on the road-sides, can be re-suspended and together with mineral particles (of CM) and traffic-related big non-exhaust particles (tire and brake lining wear, particles from corroded car parts, road surface particles, etc.) may be caught by the impactor, especially in spring, the sampling period at HW. In summer, at CR in the city center, roads and pavements are daily cleaned and sprinkled with water to limit the dust re-suspension. The winter deicing salt had been removed long before the sampling at CR started. Probably, the majority of big particles in CM at CR did not come from re-suspension, but they were brought by the wind from lawns and green belts.
Besides the winter road deicing, the most probable sources of PM1 and PM1–2.5
bound Na+ and Cl– in Katowice are domestic stoves (burning low quality coal and
household wastes [11–13]). The high contribution of NaCl to PM at CR in summer, due to high concentrations of Na+ (Table 1), may also be related with biomass burning
in the CR neighborhood [4, 6, 8].
The mass of TE in PM was also significant at the two sites. At CR, its contribu-tions to PM1 and to PM1–2.5 were twice the contributions at HW. The TE contributions
to coarse PM were similar at both sites (Table 3). The greater at CR than at HW amount of fine PM bound metals (probably from oxides) in TE is probably caused by municipal emissions – in the holiday season the traffic density decreased significantly at CR. At HW, far from the city core, the TE probable basic source was road traffic. At both sites, the coarse PM bound TE probably came from non-exhaust emissions: tire wear, brake lining, car part corrosion, road surface abrasion, etc. Also, the measur-ing points might be affected by non-ferrous metal smelters, cokeries, lead and cadmi-um refineries, cement factories, etc., releasing some of the TE components to the air [11–13].
The best fit of the reconstructed mass was received for PM1 at CR. At both sites,
the greatest UM share was in PM10–40. Water may account for up to 30% of the PM
mass in some periods of the year [22]. It seems that the water content in PM was greater at HW because of higher air humidity and lower air temperatures at this point. Such conditions favor condensation of water on the airborne particles. Moreover, the composition of PM differs between the sites, the presence of ammonium nitrate, or great amounts of chlorides (calcium chloride and its hydrates from maintaining of roads in winter) might make PM more hygroscopic at HW. UM comprises also uni-dentified elements chemically bound to the elements from TE (e.g. oxygen from the oxides of elements from TE). Moreover, PM was not analyzed for all possible com-pounds and elements. The observable contributions to the near-road TE may have the platinoids, or to CM—the lanthanides, neither considered in the work. And, of course, it should always be kept in mind that the sample taking, sampling campaign duration,
artifacts arising in the transport, weighing, chemical analyses, and many others, affect the difference between the gravimetrically and the analytically found masses of TSP.
4. CONCLUSIONS
The mass reconstruction of selected PM fractions at HW (highway) and CR (crossroads) revealed no significant differences in the PM composition between these sites and earlier investigated urban areas in Poland. At urban background sites in Po-land, at HW, and at CR alike, fine PM (PM2.5) consists mainly of carbonaceous matter
(organic matter and elemental carbon), being 46 and 52% of the PM1 and PM1–2.5
masses at HW and 61 and 57% of the PM1 and PM1–2.5 masses at CR, respectively.
Another significant component of PM2.5 is secondary inorganic aerosol (sulfates,
ni-trates, ammonium), being 19 and 18% of the PM1 and PM1–2.5 masses at HW and 21
and 15% of the PM1 and PM1–2.5 masses at CR. Coarse dust (PM2.5–40), at both sites,
consists mainly of carbonaceous matter (48 and 45% of PM2.5–10 and PM10–40 masses at
HW, 34 and 36% of PM2.5–10 and PM10–40 masses at CR) and mineral/crustal matter
(21 and 18% of PM2.5–10 and PM10–40 masses at HW, 45 and 36% of PM2.5–10 and PM10–40
masses at CR).
The chemical composition of PM at both HW and CR differs from the PM compo-sition at other traffic affected sites in Europe. Elemental carbon contributes less to PM2.5 and more to PM2.5–40, PM2.5 contains more organic carbon and significantly less
secondary inorganic aerosol in Katowice than elsewhere in Europe. It is hard to ex-plain what these differences are due to because the results of any investigation depend strongly on the sapling and analytical methods, equipment, averaging of results, calcu-lations for chemical mass closure checking, etc. However, the specificity of the chem-ical composition of PM and the PM components in southern Poland, compared to the PM composition in other European regions, seems to be beyond doubt. A very im-portant in Poland, negligible in the majority of other European countries, source of PM and PM bound elemental carbon is the municipal emissions (combustion of low-quality coal, coal dust, biomass, and household waste in domestic ovens, combustion of hard coal in local heating plants). Therefore, although it was shown here that traffic affects the formation of the secondary aerosol in PM2.5 and the shares of soil matter
and NaCl in PM2.5–40 in Katowice, the traffic effect on the concentrations and chemical
composition of PM may be significantly lower in southern Poland than elsewhere in Europe.
ACKNOWLEDGMENTS
The work was realized within the projects Nos. C.1.2 and N N523 564038, the former financed by the Institute of Environmental Engineering, PAS, the latter – by the Polish Ministry of Science and High-er Education.
REFERENCES
[1] KÜNZLI N.,KAISER R.,MEDINA S.,STUDNICKA M.,CHANEL O.,FILLIGER P.,HERRY M.,HORAK JrF., PUYBONNIEUX-TEXIER V.,QUÉNEL P.,SCHNEIDER J.,SEETHALER R.,VERGNAUD J.-C.,SOMMER H., Public-health impact of outdoor and traffic-related air pollution: a European assessment, Lancet,
2000, 356, 795.
[2] LEVY J.I.,BUONOCORE J.J.,VON STACKELBERG K., Evaluation of the public health impacts of traffic
congestion: a health risk assessment, Environ. Health, 2010, 9.
[3] PANT P.,HARRISON R.M., Estimation of the contribution of road traffic emissions to particulate
mat-ter concentrations from field measurements: A review, Atmos. Environ., 2013, 77, 78.
[4] PUTAUD J.-P.,RAES F.,VAN DINGEN R.,BRÜGGEMANN R.,FACCHINI M.-C.,DECESARI S.,FRUZZI S., GEHRIG R.,HÜGLIN R.,LAJ P.,LORBEER G.,MAENHAUT W.,MIHALOPOULOS N.,MÜLLER K.,QUEROL X.,
RODRIGUEZ S.,SCHNEIDER J.,SPINDLER G.,BRINK H.,TØRSETH K.,WIEDENSOHLER A., European
aerosol phenomenology . 2. Chemical characteristics of particulate matter at kerbside, urban, rural and background sites in Europe, Atmos. Environ., 2004, 38, 2579.
[5] PEREZ L.,MEDINA-RAMÓN M.,KÜNZLI N.,ALASTUEY A.,PEY J.,PEREZ N.,GARCIA R.,TOBIAS A.,
QUEROL X.,SUNYER J., Size fractionate particulate matter, vehicular traffic, and case-specific daily
mortality in Barcelona (Spain), Environ. Sci. Technol., 2009, 43, 4707.
[6] VIANA M., KUHLBUSCH T.A.J., QUEROL X., ALASTUEY A., HARRISON R.M., HOPKE P.K., WINIWARTER W.,VALLIUS W.,SZIDAT S.,PRÉVÔT A.S.H.,HUEGLIN C.,BLOEMEN H.,WÅHLIN P.,
VECCHI R.,MIRANDA A.,KASPER-GIEBL A.,MAENHAUT W.,HITZENBERGER R., Source
apportion-ment of particulate matter in Europe: A review of methods and results, Aerosol Sci., 2008, 39, 827.
[7] KONG S.,JI Y.,LU B.,BAI Z.,CHEN L.,HAN B.,LI Z.,Chemical compositions and sources of atmos-pheric PM10 in heating, non-heating and sand periods at a coal-based city in northeastern China, J. Environ. Monitor., 2012, 14, 852.
[8] SILLANPÄÄ M., HILLAMO R., SAARIKOSKI S.,FREY A.,PENNANEN A., MAKKONEN U.,SPOLNIK Z.,
VAN GRIEKEN R.,BRANIŠ M.,BRUNEKREEF B.,CHALBOT M.,KUHLBUSCH T.,SUNYER J.,KERMINEN V.M., KULMALA M.,SALONEN R.O., Chemical composition and mass closure of particulate matter at six urban sites in Europe, Atmos. Environ., 2006, 40, 212.
[9] YIN J.,HARRISON R.M., Pragmatic mass closure study for PM1.0, PM2.5 and PM10 at roadside, urban
background and rural sites, Atmos. Environ., 2008, 42, 980.
[10] THEODOSI C.,GRIVAS G.,ZARMPAS P.,CHALOULAKOU A.,MIHALOPOULOS N., Mass and chemical composition of size-segregated aerosols (PM1, PM2.5, PM10) over Athens, Greece: local versus re-gional sources, Atmos. Chem. Phys., 2011, 11, 11895.
[11] ROGULA-KOZŁOWSKA W.,KLEJNOWSKI K., Submicrometer aerosol in rural and urban backgrounds
in Southern Poland: Primary and secondary components of PM1, B. Environ. Contam. Tox., 2013,
90, 103.
[12] ROGULA-KOZŁOWSKA W.,KLEJNOWSKI K.,ROGULA-KOPIEC P.,MATHEWS B.,SZOPA S.,A study on the seasonal mass closure of ambient fine and coarse dusts in Zabrze, Poland, B. Environ. Contam.
Tox., 2012, 88, 722.
[13] ROGULA-KOZŁOWSKA W., BŁASZCZAK B., SZOPA S.,KLEJNOWSKI K.,SÓWKA I., ZWOŹDZIAK A., JABŁOŃSKA M.,MATHEWS B., PM2.5 in the central part of Upper Silesia, Poland: concentrations, elemental composition, and mobility of components, Environ. Monit. Assess., 2013, 185, 581.
[14] MARCAZZAN G.M.,VACCARO S.,VALLI G.,VECCHI R., Characterisation of PM10 and PM2.5 partic-ulate matter in the ambient air of Milan (Italy), Atmos. Environ., 2001, 35, 4639.
[15] WEDEPOHL K.H., The composition of the continental crust, Geochim. Cosmochim. Acta, 1995, 59,
[16] MARICQ M.M., Chemical characterization of particulate emission from diesel engines: a review,
Aerosol Sci., 2007, 38, 1079.
[17] GELLER M.D.,NTZIACHRISTOS L.,MAMAKOS A.,SAMARAS Z.,SCHMITZ D.A.,FROINES J.R.,SIOUTAS C., Physicochemical and redox characteristics of particulate matter (PM) emitted from gasoline and diesel passenger cars, Atmos. Environ., 2006, 40, 6988.
[18] THORPE A.,HARRISON R.M., Sources and properties of non-exhaust particulate matter from road
traffic: a review, Sci. Total Environ., 2008, 400, 270.
[19] ROGULA-KOZLOWSKA W.,ROGULA-KOPIEC P.,KLEJNOWSKI K.,BLASZCZYK J.,Influence of vehicular traffic on concentration and mass size distribution of two fractions of carbon in an urban area at-mospheric aerosol, Rocz. Ochr. Środ., 2013, 15, 1623.
[20] CASTRO L.M.,PIO C.A.,HARRISON R.M.,SMITH D.J.T., Carbonaceous aerosol in urban and rural European atmospheres: estimation of secondary organic carbon concentrations, Atmos. Environ.,
1999, 33, 2771.
[21] GROSJEAN D.,FRIEDLANDER S.K., Gas-particle distribution factors for organic and other pollutants
in the Los Angeles atmosphere, JAPCA, 1975, 25, 1038.
[22] TSYRO S.G.,To what extent can aerosol water explain the discrepancy between model calculated and gravimetric PM10 and PM2.5?, Atmos. Chem. Phys., 2005, 5, 515.
